Determinants of mortality and type of repair in neonates with pulmonary atresia and intact ventricular septum.

Abstract

We sought to define the prevalence of definitive end states and their determinants in children given a diagnosis of pulmonary atresia and intact ventricular septum during the neonatal period. Between 1987 and 1997, 408 neonates with pulmonary atresia and intact ventricular septum were entered into a prospective study by 33 institutions. Competing risks analysis was used to demonstrate the prevalence of 6 end states. Factors predictive of attaining each end state were identified by means of multivariable analysis with bootstrap validation. Overall survival was 77% at 1 month, 70% at 6 months, 60% at 5 years, and 58% at 15 years. Prevalence of end states 15 years after entry were as follows: 2-ventricle repair, 33%; Fontan repair, 20%; 1.5-ventricle repair, 5%; heart transplant, 2%; death before reaching definitive repair, 38%; and alive without definitive repair, 2%. Patient-related factors discriminating among end states primarily included adequacy of right-sided heart structures, degree of aberration of coronary circulation, low birth weight, and tricuspid valve regurgitation. After adjusting for these factors, 2 institutions were predictive of 2-ventricle repair, 1 of Fontan repair, and 6 of death before definitive repair. Two institutions were predictive of both 2-ventricle and Fontan repair. These 2 institutions achieved a higher risk-adjusted prevalence of definitive repair and a lower prevalence of prerepair mortality. Characteristics of neonates with pulmonary atresia and intact ventricular septum predict type of definitive repair. A morphologically driven institutional protocol emphasizing both 2-ventricle and Fontan pathways might mitigate the negative effect of unfavorable morphology. In the current era, 85% of neonates are likely to reach a definitive surgical end point, with 2-ventricle repair achieved in an estimated 50%.